• Journal of Korea Water Resources Association
• /
• v.32 no.1
• /
• pp.3-13
• /
• 1999

In this study, an alternative spatial analysis method against conventional methods such as Thiessen method, Inverse Distance method, and Kriging method, named Spatial-Analysis Neural-Network (SANN) is presented. It is based on neural network modeling and provides a nonparametric mean estimator and also estimators of high order statistics such as standard deviation and skewness. In addition, it provides a decision-making tool including an estimator of posterior probability that a spatial variable at a given point will belong to various classes representing the severity of the problem of interest and a Bayesian classifier to define the boundaries of subregions belonging to the classes. In this paper, the SANN is implemented to be used for analyzing a mean annual precipitation filed and classifying the field into dry, normal, and wet subregions. For an example, the whole area of South Korea with 39 precipitation sites is applied. Then, several useful results related with the spatial variability of mean annual precipitation on South Korea were obtained such as interpolated field, standard deviation field, and probability maps. In addition, the whole South Korea was classified with dry, normal, and wet regions.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.3
• /
• pp.188-193
• /
• 2003

Regardless of the distribution of the adaptation data in the testing environment, model-based adaptation methods that have so far been reported in various literature incorporates the adaptation data undiscriminatingly in reducing the mismatch between the training and testing environments. When the amount of data is small and the parameter tying is extensive, adaptation based on outlier data can be detrimental to the performance of the recognizer. The distribution of the adaptation data plays a critical role on the adaptation performance. In order to maximally improve the recognition rate in the testing environment using only a small number of adaptation data, supervised weighted training is applied to the structural maximum a posterior (SMAP) algorithm. We evaluate the performance of the proposed weighted SMAP (WSMAP) and SMAP on TIDIGITS corpus. The proposed WSMAP has been found to perform better for a small amount of data. The general idea of incorporating the distribution of the adaptation data is applicable to other adaptation algorithms.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.5
• /
• pp.2603-2609
• /
• 2014

Reliability analysis of mechanical systems requires statistical modeling of input random variables such as distribution function types and statistical parameters that affect the performance of the mechanical systems. Some random variables are correlated, but considered as independent variables or wrong assumptions on input random variables have been used. In this paper, joint distributions were modeled using copulas and Bayesian method from limited number of data. To verify the proposed method, statistical simulation tests were carried out for various number of samples and correlation coefficients. As a result, the Bayesian method selected the most probable copula types among candidate copulas even though the candidate copula shapes are similar for low correlations or the number of data is limited. The most probable copulas also yielded similar reliabilities with the true reliability obtained from a true copula, so that it can be concluded that the Bayesian method provides accurate statistical modeling for the reliability analysis.

• Composites Research
• /
• v.35 no.3
• /
• pp.188-195
• /
• 2022

This paper proposes a multiscale process-structural analysis methodology and applies to a battery housing part made of the short fiber-reinforced and fabric-reinforced composite layers. In particular, uncertainties of the material properties within the microscale representative volume element (RVE) were considered. The random spatial distribution of matrix properties in the microscale RVE was realized by the Karhunen-Loeve Expansion (KLE) method. Then, effective properties of the RVE reflecting on spatially varying matrix properties were obtained by the computational homogenization and mapped to a macroscale FE (finite element) model. Morever, through the hybrid process simulation, a FE (finite element) model mapping residual stress and fiber orientation from compression molding simulation is combined with one mapping fiber orientation from the draping process simulation. The proposed method is expected to rigorously evaluate the design requirements of the battery housing part and composite materials having various material configurations.

• Journal of The Geomorphological Association of Korea
• /
• v.26 no.2
• /
• pp.1-14
• /
• 2019

This study analyzed the relativity between block stream and talus distributions by employing a likelihood ratio approach. Possible distribution sites for each debris slope landform were extracted by applying a spatial integration model, in which we combined fuzzy set model, Bayesian predictive model, and logistic regression model. Moreover, to verify model performance, a success rate curve was prepared by cross-validation. The results showed that elevation, slope, curvature, topographic wetness index, geology, soil drainage, and soil depth were closely related to the debris slope landform sites. In addition, all spatial integration models displayed an accuracy of over 90%. The accuracy of the distribution potential area map of the block stream was highest in the logistic regression model (93.79%). Eventually, the accuracy of the distribution potential area map of the talus was also highest in the logistic regression model (97.02%). We expect that the present results will provide essential data and propose methodologies to improve the performance of efficient and systematic micro-landform studies. Moreover, our research will potentially help to enhance field research and topographic resource management.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2003.11a
• /
• pp.310-313
• /
• 2003

최근 들어 립리딩은 멀티모달 인터페이스 기술의 응용분야에서 많은 관심을 모으고 있다. 동적영상을 이용한 립리딩 시스템에서 해결해야 할 주된 문제점은 상황 변화에 독립적으로 얼굴 영역과 입술 영역을 추출하고 오프라인이 아닌 실시간으로 입력된 입술 영상의 인식을 처리하여 립리딩의 사용도를 높이는 것이다. 본 논문에서는 사용자가 쉽게 사용할 수 있는 PC카메라를 사용하여 영상을 입력받아 학습과 인식을 실시간으로 처리하는 립리딩 시스템을 구현하였다. 본 논문에서는 움직임이 있는 화자의 얼굴영역과 입술영역을 컬러, 조명등의 변화에 독립적으로 추출하기 위해 HSI모델을 이용하였다. 입력 영상에서 일정한 크기의 영역에 대한 색도 히스토그램 모델을 만들어 색도 영상에 적용함으로써 얼굴영역의 확률 분포를 구하였고, Mean-Shift Algorithm을 이용하여 얼굴영역의 검출과 추적을 하였다. 특징 점 추출에는 이미지 기반 방법인 PCA 기법을 이용하였고, HMM 기반 패턴 인식을 사용하여 실시간으로 실험영상데이터에 대한 학습과 인식을 수행할 수 있었다.

• Composites Research
• /
• v.14 no.2
• /
• pp.50-59
• /
• 2001

The predictions of residual strength evolution and fatigue life of full scale composite rotor blade for multipurpose helicopter were studied using a strength degradation model. Flight-by-flight load spectrum was developed on the basis of FELIX standard spectrum data. The laminated structural analysis was also performed to obtain corresponding local stress and/or strain spectra for each ply of laminate skin and glass roving spar structures around the blade root where fatigue damage was severely anticipated.

• Journal of Global Scholars of Marketing Science
• /
• v.11
• /
• pp.129-146
• /
• 2003

This study explores whether a NBD type of model can be applied to characterize the underlying frequency distribution of online consumer's visit behavior. In this study, the following two research questions are addressed: (1) How can we characterize the underlying distribution pattern(s) of the number of repeat i i visits to a site? (2) How can consumer's Internet usages and his/her demographics affect the average number of visits to the site? Through the empirical investigation, this study found that NBD models are directly applicable to characterize the underlying distribution of visit frequency on the Internet. Furthermore, this study addresses some managerial implications for understanding how site visits are determined. Especially this study highlights the relationship between repeated visits and the visitors' Internet Usages and demographics. The proposed models are estimated and validated by online panel data that covers more than 1000 different sites and has 800,000 observations.

• Journal of Soil and Groundwater Environment
• /
• v.10 no.5
• /
• pp.52-60
• /
• 2005

The conceptual model of under-sampled study area will include a great amount of uncertainty. In this study, we investigate the applicability of Markov chain model in a spatial domain as a tool for minimizing the uncertainty arose from the lack of data. A new formulation is developed to generalize the previous two-dimensional coupled Markov chain model, which has more versatility to fit any computational sequence. Furthermore, the computational algorithm is improved to utilize more conditioning information and reduce the artifacts, such as the artificial parcel inclination, caused by sequential computation. A generalized 20 coupled Markov chain (GCMC) is tested through applying a hypothetical soil map to evaluate the appropriateness as a substituting model for conventional geostatistical models. Comparing to sequential indicator model (SIS), the simulation results from GCMC shows lower entropy at the boundaries of indicators which is closer to real soil maps. For under-sampled indicators, however, GCMC under-estimates the presence of the indicators, which is a common aspect of all other geostatistical models. To improve this under-estimation, further study on data fusion (or assimilation) inclusion in the GCMC is required.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2018.05a
• /
• pp.312-312
• /
• 2018

우리나라의 기후 지형적 특성에 따라 연강수량의 50% 이상이 여름철에 내리며 이러한 짧은 기간에 집중적으로 내리는 강수패턴 조건하에서 수공구조물 설계시 대부분 극치빈도분석을 활용한다. 우리나라의 경우 단일 Gumbel 분포를 활용한 극치빈도분석을 많이 이용한다. 하지만, 최근 이상기후로 인하여 전세계적으로 강수패턴의 특징이 급격히 변하고 있으며, 우리나라의 강수패턴 또한 바뀌어가고 있다. 연강수량의 대부분은 태풍과 장마로 인한 강수량으로 이루어져 있고, 일반적으로 두 개의 모집단으로 이루어진 형태를 보인다. 앞선 연구에서 두 개 이상의 첨두를 가지는 형태의 연최대강수량 자료에 대해 8개의 지속시간별(1, 2, 3, 6, 9, 12, 18, 24hr)로 Bayesian 기법의 단일 Gumbel 분포형과 혼합 Gumbel분포형 기반의 극치빈도분석 결과를 비교하였고, 혼합 Gumbel 분포형이 이중첨두 부분의 거동을 효과적으로 모의하는 것을 확인하였다. 본 연구에서는 이상기후로 인한 강수량의 특징의 급격한 변화에 일정한 패턴이 있음을 가정하고 이중첨두의 연 최대일강수량 자료에 대해 혼합 Gumbel 분포형 기반 비정상성 빈도분석을 실시하였다. 정상성 빈도분석과의 비교를 위해 확률분포의 매개변수 산정시 우도함수를 Bayesian 기법을 통해 산정하여 각 분포형의 Bayesian information criterion(BIC) 값을 비교하였다. 비정상성일 경우의 BIC 값이 정상성일 경우 보다 작게 산정되었고, 강수패턴이 경향성을 가지는 것으로 판단할 수 있었다. 비정상성 혼합 Gumbel 분포형 모델은 최근 급격한 강수패턴의 변화에 대한 대응책으로서 활용성이 높을 것으로 기대된다.